French TV 'debunks' AI detector Pangram—using a random Google tool instead
paulnovosad · x · 2026-09-24
TV5MONDE ran an on-air segment claiming to show AI detector Pangram is unreliable, but the tests actually used a random low-quality detector found on Google—not Pangram at all. The quoted poster calls the level of incompetence "almost frightening." The original poster quips that critics conflating Pangram with any "Google AI detector" is actually good for Pangram: it only works well until it's common knowledge that it does.
More from Fun
- Dev jokes about adding 'digest context when agitated' to every agents.md — nptacek · 2026-09-24
- Users mock Dario Amodei's 'pace the frontier' plea as new Anthropic models line up — shaunralston · 2026-09-24
- "How Aligned Are You With Jev?" — the Decision Model Becomes an AI Meme — PawarBI · 2026-09-24
- ruff author jokes he can't in good faith release a CLI named "penguin" — charliermarsh · 2026-09-24
- Qwen 3.8 Flash Next plays Baldur's Gate 2 in a live agent experiment — Reasonable_Friend_77 · 2026-09-24
- Coding with AI agents: 'like working with a really, really fast idiot' — chaumian · 2026-09-24